[unreadable] [unreadable] The early diagnosis of liver cancer is critical in determining relevant treatment methods. Imaging techniques, including magnetic resonance imaging (MRI) techniques continue to be developed and while standard imaging techniques such as T2 and T1 weighted MRI lead to visualization of nodules, adenomas and tumors, it is generally too late for the patient to be successfully treated due to the limitation of detection with standard MRI. In our laboratory we have pursued the diagnosis of liver disease by observing the profiles of biochemical markers using nuclear magnetic resonance (NMR) spectra. Using a customized radio-frequency pulse program based on single quantum coherence spectroscopy, our preliminary year long study of hepatocarcinogenesis in the rat, induced by a choline deficient (CD) diet, showed marked differences in fatty acid species when compared to controls. Such a pulse sequence, wherein tailored and selective radio-frequency pulses were incorporated within the 2D heteronuclear correlation (HSQC) framework, allowed for the unequivocal determination of the number of double bonds contained in an unsaturated fatty acid containing compound. Previously, the signal from NMR spectra of these compounds were thought to be chemically equivalent and thus could not be resolved to implicate the presence of an individual fatty acid moiety. However, by using this pulse sequence, we have found that in CD animals, the fatty acid profile within weeks of being on the diet, is dominated by compounds containing two to four double bonds and with age is dominated by compounds containing one double bond at a concentration that is approximately four times greater than the CS (choline sufficient) controls. We propose here that the method of selectively observing chosen signals in an in vivo spectrum, will be a valuable method for characterizing the state of the liver at a very early stage. Improvements to the concepts, methods and pulse sequence programs leading to the tailored HSQC sequence mentioned above will lead to development of novel pulse sequences specifically designed to detect individual metabolites, and will lead to improved clinical MR methods suitable for accurate and early diagnosis of disease. This forms the basis for one of two specific aims of the proposed course of study. In the second part of our study we propose to implement the methods on a small animal magnetic resonance imaging spectrometer and test the validity and accuracy of the methods on a transgenic mouse model of liver cancer at chosen time-points over a period of 8 months. The results from this study will be compared with normal controls. At the end of each time point, the liver from each group of mice will be excised and fatty acid speciation determined by our modified HSQC method and cross-validated with the results obtained by the customized pulse sequence methods deployed on the small MRI spectrometer. While the application here stops at the animal model level, the long-term objective is to validate a non-invasive, method for the diagnosis of liver cancer by MRI/S in the clinic and one that augments current staging methods. [unreadable] [unreadable] [unreadable] [unreadable]